Post on 09-Sep-2018
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
A Standard Qualification Plan for Composite Material Systems
M R L Gower and G D SimsDivision of Engineering and Process Control
National Physical Laboratory
Reducing qualification costs in the composite materials industry
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Content of presentation
1. Introduction
2. Formulation of Standard Qualification Plan (SQP)
3. Round-robin validation of test methods
4. Results analysis procedure
5. Finalisation of SQP
6. Future work
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
1- Introduction
What is the Standard Qualification Plan?
Standard Qualification Plan (SQP): a set of composite
material test standards that will meet the minimum
common requirements necessary to allow:
• Quality control
• Initial material selection
• Preliminary design
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Why Is a Qualification Plan Needed?
Costly to
….qualify a product against different specifications
Material A
DIN
BS
ISO
ASTM
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Why Is a Qualification Plan Needed?
Costly to
….qualify a product against different specifications
Material A
DIN
BS
ISO
ASTM
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Why Is a Qualification Plan Needed?
Costly to
….introduce new materials because of qualification costs
Material A
Material B
Material C
ApplicationX
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Why Is a Qualification Plan Needed?
Difficult to
….find data for materials selection and preliminary design
Material A
Material B
Material C
LACK OF
DATAX
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
• SQP feasible due to recent availability of:
– test panel manufacture standard (ISO 1268)
– suite of harmonised test methods (mechanical, thermal and physical – see http://www.npl.co.uk/npl/cmmt/cog/index.html)
– a data-sheet database standard (ISO 10350-2)
• New composite materials could be released with qualification data
• Previous consultation with industry indicated considerable support
Feasibility
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Beneficiaries
• End-users/designers
• Suppliers of prepreg composite materials
• Certification bodies e.g. CAA, FAA etc.
• Test houses
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Project Objectives
1. Develop a Standard Qualification Plan (SQP) aimed at significantly reducing qualification costs
2. Demonstrate to industry, the suitability and robustnessof test methods proposed in the SQP
3. Recommend and disseminate to industry
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
2 Formulation Process
• Extensive industry consultation to agree content of plan– tests most commonly required/performed
– test standards being used
– use of test data
– importance of individual test methods and the need for their inclusion in the SQP
• Output - a draft report with an industrial feedback mechanism
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Test MethodsUNCURED• Mass per unit area• Resin flow• Fibre mass per unit area• Percentage of volatile matter• Resin and fibre volume fraction• Glass transition temperature• Analysis by DSC• Density of fibre• Density of resin• Gell time
CURED• Tension – unidirectional• Tension - multidirectional• Compression – unidirectional• Compression – multidirectional• Shear ± 45° tension
• Shear strength – In-plane• ILSS – Through-thickness• Flexural• Mode I fracture toughness• Mode II fracture toughness• Filled/open hole tension• Filled/open hole compression• Pin-bearing (plain, un-torqued)• Bolted-joint bearing (torqued)• Compression-after-impact• Fatigue• Creep• Coefficient of thermal
expansion• Moisture uptake/conditioning• Effect of water/moisture• Effect of chemicals• Effect of heat ageing
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Questionnaire results
• uncured properties required for materials selection and quality assurance purposes
• uncured property test methods - medium importance, much of this data covered by material suppliers anyway.
• cured properties used mainly for materials selection and design
• cured property test methods - high importance, especially:– Tension– Compression– Shear– ILSS through-thickness
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
SQP/EQP format
• Draft SQP/EQP written in the style of an ISO standard• Two parts
– Part A – rationale and instructions– Part B – report sheets
• Based on format of– BS EN ISO 11403-1 – Acquisition and presentation of
comparable multi-point data– ISO 10350-2 - Acquisition and presentation of comparable
single-point data
• ISO test standards chosen as default• Data included for several ISO test temperatures after
dry and hot/wet conditioning
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Specimen Sampling
• Each property (per test environment) to be determined from a series of 30 tests
• 3 material batches– 2 panels per batch
– 5 specimens per panel
• MIL-HDBK-17 recommends a min. of 30 specimens taken from at least 5 batches
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Specimen Sampling
BATCH 1
PANEL 1 PANEL 2
5 SPECIMENS 5 SPECIMENS
BATCH 2
PANEL 3 PANEL 4
5 SPECIMENS 5 SPECIMENS
BATCH 3
PANEL 5 PANEL 6
5 SPECIMENS 5 SPECIMENS
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Presentation of results• Results presented in report sheets of Part B
– mean and standard deviations of 30 tests– individual results provided to allow further data analysis– procedure provided for generation of A- and B- basis design allowables –
based on methods of MIL-HDBK-17
• Certain data values recommended to be normalised with respect to a nominal Vf – procedure provided
• Normalise all mechanical stiffness and strength properties except:− 90° (transverse) tension (UD laminates)− 90° compression (UD laminates)− interlaminar shear− in-plane shear− short beam strength− bearing− strain energy release rates− Poisson’s ratio
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
• Round-robin (RR) conducted on 6 test methods
• Tests selected for RR dependent on importance of data, likelihood of error in testing, availability of previous data
• Analysis of results to ISO 5725
• Organisations were encouraged to assess at an early stage the use of the SQP in their operations
3 Round-robin validation of test methods
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Materials
T700 (24k)SE84LV/HSC/300/300/37±3%SP Systems2
T300J (12k)913 Carbon-T300J-5-35%Hexcel1
Fibre typeDescriptionSupplierMaterial
N.B. Both materials donated by industry
Typical of Aerospace/Formula 1 automotive applications
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Panel/specimen preparation
• Unidirectional specimens machined from 1 and 2 mm thick panels
• Test panels manufactured according to– ISO 1268 Part 4 – Preparation of fibre-reinforced, resin bonded, low-
pressure, laminated plates or panels for test purposes
• Specimens extracted following – ISO 2818 – Preparation of test specimens by machining
• Additional machining guidance– Measurement Good Practice Guide - Machining of Composites and
Specimen Preparation (NPL GPG No. 38)
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Test Methods
ISO/CD 6721-11
BS EN ISO 14130
BS EN ISO 14125BS EN ISO 14126BS EN ISO 527-1 and -5
Standard
Tg, Tonset, Tloss, TtandeltaDMA
τm1Interlaminar shear (ILSS)
σmf11, Ef11Flexureσmc11, Ec11Compressionσmt11, Et11, ν12TensionProperties measuredTest
Mechanical tests – 6 specimens, DMA – 3 specimens
Specimen preparation undertaken by NPL
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Tensile tests to BS EN ISO 527-5
250 mm
15 mm50 mm Thickness = ~ 1 mm
• Biaxial strain gauge• Dynamic serial gauge• Biaxial extensometer• Longitudinal and transverse extensometers• Crosshead deflection
Method of deflection measurement
100, 250Load cell capacities (kN)
2 used by allTest speeds used (mm/min)
8Number of sites
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Compression tests to BS EN ISO 14126
tab
material110 mm
10 mm10 mm
• Celanese• In-house end loading blocks
Loading jig
• Biaxial strain gauges• Extensometer• Crosshead deflection
Method of deflection measurement
100, 200, 250Load cell capacities (kN)
1 and 1.27Test speeds used (mm/min)
5Number of sites
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Flexure tests to BS EN ISO 14125
• 7 sites used different span for each material • 2 sites used one span for both materials
Span (3-point loading for all sites)
• LVDT• Crosshead deflection
Method of deflection measurement
1, 5, 10, 20, 50, 100Load cell capacities (kN)
1, 2, 5, 5.21, 6.6, 6.75, 7(3 sites used different test speeds for each material)
Test speeds used (mm/min)
9Number of sites
ILSS tests to BS EN ISO 14130
All sites used different span for each materialSpan5, 100Load cell capacities (kN)
1 used by allTest speeds used (mm/min)
6Number of sites
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
DMA tests to ISO/CD 6721-11
25-200, 25-250, 25-300Temperature range (°C)
1 Hz used by allFrequency
3Heating rate (°C/min)
• 20 mm for 3-point bend• 10, 15, 17.5 mm for single cantilever bend• 20 mm for dual cantilever bend
Test span (range)
• 1 site used a 3-point bend configuration• 3 sites used a single cantilever bend configuration• 1 site used single and dual cantilever configurations
Test mode
5Number of sites
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Participants
• Advanced Composites Group• Motor Industry Research Association (MIRA)• Ford Motor Company• Slingsby Aviation• Aerostructures Hamble• Composites Testing Laboratory• Bureau Veritas• Gearing Scientific• Triton Technology • Perkin Elmer Thermal Analysis Solutions• NPL
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
12568935488Total
1110987654321
τM1Ef11σMf11Ec11σMc11ν12Et11σMt11DMA
ILSSFlexureCompressionTensionSite
Properties measured
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
• ISO 5725-2 - Accuracy (trueness and precision) of measurement methods and results -- Part 2: Basic method for the determination of repeatability and reproducibility of a standard measurement method
4 Analysis of results
Initial check of “as-received”
data Mandel’s h and k statistics
Cochran and Grubb tests for
outliers and stragglers
Calculation of repeatability
and reproducibility
III
IIIIV
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
I - Check of “as-received” data
0
50
100
150
200
250
0 1 2 3 4 5 6 7 8 9
Site
E t11
(GPa
)
1500
2000
2500
3000
0 1 2 3 4 5 6 7 8 9
Site
σMt1
1 (M
Pa)
0
200
400
600
800
1000
1200
1400
0 1 2 3 4 5 6 7 8
Site
σMc1
1 (M
Pa)
Tensile strength Tensile modulus
Compression strength
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
II – Mandel’s h and k consistency statistics
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
1 3 4 5 6 7 9
Site
Bet
wee
n si
te c
onsi
sten
cy, h
ij
Material 1 Material 2
0
0.5
1
1.5
2
2.5
1 3 4 5 6 7 9
Site
With
in si
te c
onsi
sten
cy, k
ij
Material 1 Material 2
Tensile modulus -example of normal
pattern
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
II – Mandel’s h and k consistency statistics
Flexural modulus -example of abnormal
pattern
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
1 2 3 4 5 6 7 9
Site
Bet
wee
n si
te c
onsi
sten
cy, h
ij
Material 1 Material 2
0
0.5
1
1.5
2
2.5
3
1 2 3 4 5 6 7 9
Site
With
in si
te c
onsi
site
ncy,
kij
Material 1 Material 2
0
2040
60
80
100120
140
160180
200
0 1 2 3 4 5 6 7 8 9Site
Ef11
(GPa
)
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
III – Cochran and Grubb tests
• Cochran test – site variances
• Grubb test – site means and outlying observations
• Following ISO 5725:– Outliers discarded– Stragglers retained
• On basis of statistical tests some further data discarded
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Tension results
0
500
1000
1500
2000
2500
3000
1 2 3 4 5 6 7 9 S
Site
σMt1
1 (M
Pa)
Material 1 Material 2
0
20
40
60
80
100
120
140
160
1 3 4 5 6 7 9 SSite
E t11
(GPa
)
Material 1 Material 2
0
0.1
0.2
0.3
0.4
1 3 5 9
Site
ν12
Material 1 Material 2
• Explosive fracture
• Various deflection measurement methods used
• Good repeatability and reproducibility after removal of erroneous data
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Compression results
0
200
400
600
800
1000
1200
1400
1600
1 3 5 7 S
Site
Mc1
1 (M
Pa)
Material 1 Material 2
0
20
40
60
80
100
120
140
1 3 5 S
Site
Ec11
(GPa
)
Material 1 Material 2
• Acceptable failure modes achieved
• Only 1 site checked for bending
• High values of repeatability and reproducibility for strength
• Few sites able to undertake tests
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Flexure results• Acceptable failure modes
achieved
• Various deflection measurement methods used
• Various test speeds used
• Systematic errors in measurement observed for 2 sites
• Good repeatability and reproducibility after removal of erroneous data
0
500
1000
1500
2000
2500
1 2 3 4 5 6 7 8 9
Site
σMf1
1 (M
Pa)
Material 1 Material 2
0
20
40
60
80
100
120
140
160
180
1 2 3 4 5 6 7 9
Site
E f11
(GPa
)
Material 1 Material 2
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
ILSS results
0
20
40
60
80
100
120
1 3 5 7 8 9 SSite
τM11
(MPa
)
Material 1 Material 2
• Failure modes unacceptable for all sites and both materials
• Data analysed as purpose of this round-robin was not to generate precision data
• Good repeatability and reproducibility after removal of erroneous data
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
DMA results
0
20
40
60
80
100
120
140
160
180
200
3 5 10 11 12Site
Tem
pera
ture
, °C
Tg Tonset Tloss Ttandelta
020
4060
80100
120140
160180
200
3 5 10 11 12Site
Tem
pera
ture
, °C
Tg Tonset Tloss Ttandelta
Material 1
Material 2
• Double peaks reported on Ttandeltaplots for material 1 – not fully cured
• Some difficulties specifying onset and loss modulus peaks
• very low repeatability
• higher reproducibility due to
- deficiencies in temperature measurement
- various methods for temperature calibration
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
0
5
10
15
20
sMt1
1
Et11 n1
2
sMc1
1
Ec11
sMf1
1
Ef11
tM1 Tg
Tons
et
Tlos
s
Ttan
delta
Perc
enta
ge o
f gen
eral
mea
n (%
)Repeatability CoV - Material 1 Repeatability CoV - Material 2Reproducibility CoV - Material 1 Reproducibility CoV - Material 2
σMt11 Et11 ν12 σMc11 Ec11 σMf11 Ef11 τM1 Τg Τonset Τloss Τtandelta
TENSION COMPRESSION FLEXURE ILSS DMA
IV – Repeatability and reproducibility
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
Conclusions
• Tensile, flexure, ILSS and compression modulus tests showed good repeatability and reproducibility values
• Compression strength showed considerable within and between site variability – alignment crucial (essential to check for bending)
• Further development and guidance needed for compression testing aspects such as strain measurement, end-tab design and testing of thick sections
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
• Improvements required in DMA temperature measurement and calibration– NPL has developed 3 types of calibration specimen – one of which
will be used in a DMA calibration standard (ISO/CD 6721-12)
• Use of accurate, calibrated equipment essential for measuring :– Load– Displacement– Specimen dimensions (to accuracy required by standard)
• Displacement/strain should where possible be measured using:– Clip gauge extensometers– Strain gauges– LVDT/dial gauge indicators
• At the very least the crosshead displacement should only be used in conjunction with a machine compliance correction
Conclusions
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
• Finalisation of SQP as a Good Practice Guide based on feedback from draft and round-robin exercise
• Used as a pre-cursor for standardisation
• Promoted widely to the composites industry
• Assessment of current status of material databases
5 Finalisation of SQP
COMPTEST 2004, 21-23 SeptemberUniversity of Bristol
6 Future work
Further effort required for development toward standardisation
Further collaboration with CAA
Collaboration with similar international initiatives (e.g. FAA)
Development of SQPs for other materials/sectors
Additional dissemination to promote and encourage adoption